Research map / Architectures and efficiency / Spiking networks
Neural Dynamics Self-Attention for Spiking Transformers
Dehao Zhang, Fukai Guo, Shuai Wang, Jingya Wang, Jieyuan Zhang, Yimeng Shan, Malu Zhang, Yang Yang, Haizhou Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add LRF-SSA | Spikformer-8-512 · ImageNet-1K | top-1 accuracy: 73.38 -> 74.62 | Improved |
| add LRF-SSA | Spikformer-8-768 · ImageNet-1K | top-1 accuracy: 74.81 -> 75.66 | Improved |
| add LRF-SSA | QKFormer HST-10-384 · ImageNet-1K | top-1 accuracy: 78.80 -> 79.24 | Improved |
| add LRF-SSA | QKFormer HST-10-512 · ImageNet-1K | top-1 accuracy: 82.04 -> 82.52 | Improved |
| add LRF-SSA | SDT-V3 Efficient-Transformer-S · ImageNet-1K | top-1 accuracy: 75.30 -> 76.22 | Improved |
| add LRF-SSA | SDT-V3 Efficient-Transformer-L · ImageNet-1K | top-1 accuracy: 79.80 -> 80.31 | Improved |
| replace SSA with LRF-Dyn | SDT-V3 Efficient-Transformer-S · ImageNet-1K | top-1 accuracy: 75.30 -> 76.12 | Improved |
| replace SSA with LRF-Dyn | Spikformer-8-512 · ImageNet-1K | inference memory: 49.4% memory reduction with +1.13% accuracy | Improved |
| add LRF-SSA | SDT-V3 · ADE20K | mIoU: 33.6 -> 36.2 | Improved |
| replace SSA with LRF-Dyn | SDT-V3 · ADE20K | mIoU: 33.6 -> 36.3 | Improved |
| increase LRF kernel radius from none to 5 (ablation) | LRF-SSA on Spikformer · CIFAR-100 | accuracy: 77.86 -> 78.64 | Improved |
| increase LRF kernel radius from none to 5 (ablation) | LRF-Dyn on Spikformer · CIFAR-100 | accuracy: 77.78 -> 78.57 | Improved |
Closest papers by meaning
- Spiking Transformer: Introducing Accurate Addition-Only Spiking Self-Attention for Transformer (CVPR 2025 · similarity 0.92)
- SpikingResformer: Bridging ResNet and Vision Transformer in Spiking Neural Networks (CVPR 2024 · similarity 0.91)
- Spiking Transformer with Spatial-Temporal Attention (CVPR 2025 · similarity 0.90)
- Spiking Vision Transformer with Saccadic Attention (ICLR 2025 · similarity 0.90)
- STEP: A Unified Spiking Transformer Evaluation Platform for Fair and Reproducible Benchmarking (NeurIPS 2025 · similarity 0.89)
- STAA-SNN: Spatial-Temporal Attention Aggregator for Spiking Neural Networks (CVPR 2025 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
